| read.param {lazy.irtx} | R Documentation |
Reading Parameter File
read.param(infile, skip = 0, nrows = -1, na.strings = "NA", sep = "", print = 0)
infile |
Input file name |
skip |
# of lies to skip |
nrows |
NOT used |
na.strings |
A character indicating missing value |
sep |
Delimite character |
print |
= 1 to print result |
It is assumed that the data file has the following quantities
in the following order.
item_name, item_type, ncat, p1, p2, ..., p[ncat+1]
The first line must consist of the variable names shown above.
If the data file has header lines berore variable names, those lines
must be skipped by specifying skip= parameter.
When item_type == "B", ( ncat == 2 )
p1 is the discrimination parameter
p2 is the difficulty parameter
p3 is the gussing parameter
When item_type == "G",
p1 is the discrimination parameter
p2 is the category threshold parameter b_j1
p3 is the category threshold parameter b_j2
p_ncat is the category threshold parameter b_j,[ncat-1]
When item_type == "P",
p1 is the discrimination parameter
p2 is the step parameter b_j1
p3 is the step parameter b_j2
p_ncat is the step parameter b_j,[ncat-1]
When item_type == "PN,
p1 is the slope parameter
p2 is the intercept parameter b_j1
p3 is the intercept parameter b_j2
p_ncat is the intercept parameter b_j,[ncat-1]
When item_type == "N,
p1 - p_(ncat[j]-1) are the slope parameters
p_ncat[j] - p_2*(ncat[j]-1) are the intercept parameters
Regardless of the item_type, there will be ncat[j]
item paramters for item j, except for the Binary Items
which have the gusseing param as the ncat[j]+1st,
and the Nominal Items which have 2*(ncat[j]-1) item parameters.